issues raised when Idan tried to run the tests

[idanyani]

• I suggest to chmod a+x build.py
• I suggest to add "sudo apt-get install zlib1g-dev" to the SETUP_README (it seems that qemu requires it but it doesn't appear there).
• I indeed needed to install libglib2.0-dev, libpixman-1-dev (as already denoted in the SETUP_README).
• I had to search for the snapshot name in the other README file (and found it is ready_for_memory_tracer). Perhaps we should add it to the SETUP_README as well?

Finally, I tried to run the tests and encountered: simple_analysis.c: In function ‘main’: simple_analysis.c:117:5: error: ‘memset’ used with length equal to number of elements without multiplication by element size [-Werror=memset-elt-size] memset(counter_arr, 0, sizeof(counter_arr[0]) * OUR_ARR_LEN); ^~ cc1: all warnings being treated as errors

I guess it's a GCC problem since I'm using version 7.3.0 (the default for Ubuntu 18.04 LTS).

[orenmn]

fixed the first 3 (the bullets ones). good points, thanks. the last one is not a GCC problem. it's just that GCC found a bug i wrote (now fixed, hopefully). and my GCC is 6.5, so he wasn't smart enough to find it, i guess.

[idanyani]

Great Thanks! now it proceeds a bit more and then fails:

--------start running tests-------- ---start running test_analysis_tool_cmd_args--- executing cmd (in .): ['/home/idanyani/repos/qemu_mem_tracer/memory_tracer.py', '/home/idanyani/Downloads/oren_vm_disk2_final', 'ready_for_memory_tracer', '/home/idanyani/repos/qemu_with_GMBEOO', '--analysis_tool_path', '/home/idanyani/repos/qemu_mem_tracer/tests/toy_workloads_and_analysis_tools/tests_bin/simple_analysis', '--dont_add_communications_with_host_to_workload', '--workload_path_on_host', '/home/idanyani/repos/qemu_mem_tracer/tests/toy_workloads_and_analysis_tools/tests_bin/dummy_workload_with_funny_test_info', '--verbose'] executing cmd (in .): ('cat', '/proc/sys/fs/pipe-max-size') change /tmp/tmpoxkjfhpo/trace_fifo to size 1048576 (/proc/sys/fs/pipe-max-size) executing cmd (in /tmp/tmpoxkjfhpo): ['/home/idanyani/repos/qemu_mem_tracer/communications/run_qemu_and_workload.sh', '/home/idanyani/Downloads/oren_vm_disk2_final', 'ready_for_memory_tracer', '/tmp/tmpoxkjfhpo/executable1', '/home/idanyani/repos/qemu_mem_tracer/tests/toy_workloads_and_analysis_tools/tests_bin/dummy_workload_with_funny_test_info', '/home/idanyani/repos/qemu_mem_tracer/communications/write_executables_to_serial.py', 'off', '0', '0', '/home/idanyani/repos/qemu_mem_tracer/tests/toy_workloads_and_analysis_tools/tests_bin/simple_analysis', '/tmp/tmpoxkjfhpo/trace_fifo', '/home/idanyani/repos/qemu_with_GMBEOO', 'True', 'False', 'False', 'False', 'True', 'False', '2560'] ---start run_qemu_and_workload.sh--- ---starting qemu--- ---parsing qemu's message about the pseudo-terminal that it opened--- it seems that qemu terminated unexpectedly.

And a minor fix: you can now remove the following instruction from the SETUP_README:

• chmod u+x qemu_mem_tracer/build.py

[orenmn]

I added a debug print that would hopefully help understanding the cause of this failure (and similar failures caused by qemu terminating unexpectedly).

And removed the redundant instruction from SETUP_README.